Mailing-List: contact cygwin-help@sourceware.cygnus.com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe@sources.redhat.com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin@sources.redhat.com>
List-Help: <mailto:cygwin-help@sources.redhat.com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner@sources.redhat.com
Delivered-To: mailing list cygwin@sources.redhat.com
From: Chris Faylor <cgf@cygnus.com>
Date: Sat, 30 Sep 2000 22:04:25 -0400
To: cygwin@sources.redhat.com
Subject: Re: Console output oddity
Message-ID: <20000930220425.A3707@cygnus.com>
Reply-To: cygwin@sources.redhat.com
Mail-Followup-To: cygwin@sources.redhat.com
References: <i3S2hIAYIl05EwrY@gaast.demon.co.uk> <OE133tYwlib4g3Sv0YK00002446@hotmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.6i
In-Reply-To: <OE133tYwlib4g3Sv0YK00002446@hotmail.com>; from bones0_list@hotmail.com on Sun, Oct 01, 2000 at 03:17:15AM +0200

On Sun, Oct 01, 2000 at 03:17:15AM +0200, Jonas Jensen wrote:
>>I have noticed a very odd behaviour when I am sending a large amount of
>>output to the console screen.
>>If I "cat" a large text file in a bash window it takes about 20 times
>>longer than if I "type" the same file in a DOS window.  If I press and
>>hold down the CONTROL, SHIFT or ALT key as soon as the "cat" is started
>>then it runs about 4 times as fast.  (If I hold the key down while
>>starting the "cat" it runs at the lower speed but the speed increases
>>if I release the key and then press it again.) A simple C program which
>>just loops printing a counter to stdout can be speeded up by a factor
>>of about 4 by holding down a key while it runs.  It can be speeded up
>>by a factor of about 20 by compiling it with the "-mno-cygwin" switch.
>>Can anyone offer an explanation?
>
>I reported this problem some time ago, I think the topic was called
>"printf is faster when piped through tee".  I can't reproduce the
>effects you talk about when holding a key, on my system (W2k) cat is
>actually marginally faster than type.  The C program that you can speed
>up with -mno-cygwin can get almost the same speed increase if piped
>through cat or tee.
>
>AFAIK, the problem was never fixed -- "Patches are Gratefully
>ConsideredT", but I can't fix it and nobody else seems to care...

I would certainly care to have any patch that speeds up any aspect of
Cygwin.

However, if you are referring to slowness when using the CYGWIN=tty
setting, then you should notice considerable speedups in recent cygwin
snapshots.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

